Content is often displayed on devices using an aspect ratio (e.g., the ratio of the width to the height of the image) which is different than the aspect ratio of the source video. For example, 4:3 standard definition video is commonly displayed on high definition displays using a 16:9 aspect ratio. Similarly, 16:9 high definition video is commonly displayed on standard definition displays using a 4:3 aspect ratio. In such cases, aspect ratio conversion must be performed on the source video before display.
A first conventional solution for aspect ratio conversion involves scaling the source video and placing black bars on the sides or top and bottom of the scaled source video. Placing black bars on the sides of the scaled source video is commonly referred to as “pillarbox” and is used when the aspect ratio of the source video is smaller than the aspect ratio of the destination display. Placing black bars on the top and bottom of the source video is commonly referred to as “letterbox” and is used when the aspect ratio of the source video is larger than the aspect ratio of the destination display. Linear scaling, e.g., using a constant scale across the image in the horizontal direction or vertical direction, is often used to scale the source video prior to application of the black bars. Although this first conventional solution does not distort the source video, it is less desirable than other solutions since portions of the destination display (e.g., displaying the black bars) are not used and because the black bars are objectionable to some viewers.
A second conventional solution involves scaling the source video and cropping either the top and bottom of the scaled source video (e.g., when the aspect ratio of the source video is smaller than the aspect ratio of the destination display) or the sides of the source video (e.g., when the aspect ratio of the source video is larger than the aspect ratio of the destination display). In this case, the source video is scaled (e.g., enlarged or reduced) in both the horizontal and vertical directions to maintain the aspect ratio before cropping the scaled image to fit the aspect ratio of the destination display. Linear scaling is often used to scale the source video.
Although the second conventional solution utilizes more of the destination display than the first conventional solution by not applying black bars to the source video, content is lost by cropping the image. Accordingly, the second conventional solution offers a less desirable user experience since less content is displayed for viewing by a user.
A third conventional solution for aspect ratio conversion involves using linear scaling to scale the source video to fit the destination display without the use of black bars. Although the third conventional solution does not involve cropping content like the second conventional solution, the source video is distorted by the scaling of the image performed to obviate the need for the black bars. The distortion is often unacceptable to users, and therefore, provides a poor viewing experience in certain situations.
A fourth conventional solution for aspect ratio conversion involves using “parabolic scaling” to scale the source video to fit the destination display without the use of black bars. Parabolic scaling involves the use of a scaling magnitude which varies as a parabolic function across the image, where there is zero scaling at the center of the image. Accordingly, parabolic scaling distorts (e.g., squeezes or stretches) the edges of the source video more than the center of the source video (e.g., with no or little scaling). Therefore, the distortion caused by the parabolic scaling may be objectionable where important content is present near the sides of the source video, and thus, provides a poor viewing experience in certain situations.